Student opportunities

I periodically host student researchers (thesis projects, internships, working students) in egocentric vision, multimodal/embodied AI, continual learning, and industrial automation.

If you’re interested, reach out via LinkedIn or use the contact form.

Please contact me to discuss current availability and supervision.

Research topics

1) AI-Driven Robotics for Disassembly by Learning from Human Demonstration

Type: Internship or Bachelor/Master Thesis (Berlin, on-site)
Focus: Carry-forward of our egocentric research to robot policy learning for disassembly tasks (teleop/kinesthetic demos, action segmentation, policy learning, and evaluation on real hardware).

You’re a fit if you:

• Know Python + PyTorch, basic CV/ML, have ROS or robotics experience.
• Can run experiments, keep logs, and write clearly

How to apply: send a short note via Contact or message me on LinkedIn with:

• 1-page CV (PDF)
• Availability (start date, hours/week)
• Links (GitHub/portfolio)

2) Video Understanding from Egocentric & Exocentric Perspectives for Industrial Scenarios

Type: Internship or Bachelor/Master Thesis (Berlin, on-site)
Focus: Multiview video understanding in industrial workflows: activity recognition, temporal grounding, multi-modal fusion, continual learning under constraints.

You’re a fit if you:

• Comfortable with Python + PyTorch; prior CV coursework/projects
• Have an understanding of Vision-Language models (their strengths and weaknesses)
• Enjoy working with real-world datasets/pipelines and evaluating models rigorously

How to apply: use Contact or LinkedIn with the same materials as above.

General Notes

• Enrollment in a university program (BSc/MSc)
• We value initiative, reproducibility, and clear writing
• Visa/contract details depend on institute policy and role

If no roles are listed above (or no roles that suit you), feel free to reach out anyway with a short proposal (what you want to work on and why).